Technical Field
The present disclosure relates to a device for avoiding hard switching of converters, in particular in resonant converters, and a related method.
Description of the Related Art
Resonant converters are known in the state of the art, using half-bridge or full-bridge circuit topologies. In the case of a half-bridge resonant converter, the switching elements comprise a high-side transistor and a low-side transistor connected in series between an input voltage and ground. A square wave having a high value corresponding to the input voltage and a low value corresponding to ground may be generated by conveniently switching the two transistors. A small time interval Td called “dead time”, during which the transistors are turned off, is typically added immediately after each of them is turned off.
In resonant converters, the square wave generated by the half-bridge is applied to the primary winding of a transformer by a resonant network which comprises at least one capacitor and one inductor; the secondary winding of the transformer is connected with a rectifier circuit and a filter to provide a DC output voltage. The value of the output voltage depends on the frequency of the square wave.
The so-called LLC resonant converter is often used among the several types of resonant converters, especially the half-bridge LLC resonant convertor. The LLC designation comes from the resonant circuit employing two inductors (L) and a capacitor (C) and a schematic circuit of an LLC resonant converter is shown in FIG. 1. The resonant converter 1 comprises a half-bridge of transistors Q1 and Q2, with respective body diodes Db1 and Db2, between the input voltage Vin and ground GND and driven by a driver circuit 3 by means of the signals HSGD and LSGD. The common terminal HB between transistors Q1 and Q2 is connected to a resonant circuit 2 comprising a series of a capacitor Cr, an inductance Ls, and a parallel circuit that includes another inductance Lp connected in parallel to a primary of a transformer 10 with a center-tap secondary. The two windings of the center-tap secondary of transformer 10 are connected to the anodes of two diodes D1 and D2, the cathodes of which are both connected to the parallel of a capacitor Cout and a resistance Rout. The output voltage Vout of the resonant converter is the voltage across said parallel, while the output current Iout flows through the resistance Rout.
Resonant converters offer considerable advantages as compared to traditional switching converters (non-resonant, typically PWM-controlled (Pulse Width Modulation)): waveforms without steep edges, low switching losses in the power switches due to “soft” switching thereof, high conversion efficiency (>95% is easily reachable), ability to operate at high frequencies, low EMI (electro-magnetic interference) generation and, finally, high power density (i.e., enabling to build conversion systems capable of handling considerable powers levels in a relatively small space).
However, the same resonant converters are affected by certain disadvantages during the start-up step. In said step, when the high-side transistor Q1 is turned on the first time, the voltage seen by the primary winding is substantially equal to the power supply voltage. In the successive semi-period of the square wave, when the low-side transistor Q2 is turned on, the voltage seen by the primary winding is substantially equal to the voltage across the capacitor Cr; therefore, the current flowing through the resonant network increases more quickly during the turning on of the high-side transistor, while decreases less quickly during the turning on of the low-side transistor. Thereby, when the low-side transistor is turned off again, the current flows through the body diode Db2 thereof. When the high-side transistor is turned on again, a reverse voltage is developed across the body diode Db2 of the low-side transistor, while the diode Db2 is still conducting. Under said conditions, the high-side transistor is turned on under hard switching conditions and the diode Db2 is stressed in reverse recovery. Therefore, both the high-side transistor and the low-side transistor are conductive in the same time period by short-circuiting the supply terminal with the ground terminal until the body diode Db2 is recovered. Under such conditions, the voltage at the terminals of the transistor may vary so quickly that the intrinsic, parasitic bipolar transistor of the MOSFET transistor structure may be triggered, thus causing a shoot-through condition which may cause the destruction of the transistor in few microseconds.
In driving devices of high-voltage half-bridges, the power supply voltage of the driving section of the high-side MOSFET Q1 is typically obtained by means of a so-called bootstrap system, shown in FIG. 2. According to this method, the capacitor Cboot (bootstrap capacitor), is coupled with the middle point HB of the half-bridge and acts as power buffer to supply the driver 31, i.e., the part of driver 3 which drives the high-side transistor Q1. The capacitor Cboot is charged by a low-voltage generator Vcc through a high-voltage diode Dboot (bootstrap diode) with a voltage Vboot when the middle point HB of the half-bridge is at a low voltage level (that is, when the low-side transistor Q2 is turned on). When the high-side MOSFET Q1 is turned on and the middle point HB of the half-bridge is high, the diode Dboot isolates the capacitor Cboot from the low-voltage line.
Hence, to correctly drive the high-side MOSFET Q1 from the first turning-on cycle, the half-bridge is started by first turning on the low-side MOSFET Q2 so as to pre-charge the bootstrap capacitor Cboot.
In certain cases, the bootstrap diode Dboot may be provided by an integrated structure inside the driver device 3, as shown in FIG. 3. In this case, indeed, the component acting as the diode is a MOSFET transistor M, which is synchronously driven with the low-side MOSFET Q1, so as to obtain the above-mentioned functionality.
As compared to a real diode (one of ultrafast type would be used), the integrated bootstrap diode has a considerably higher resistance (of a hundred ohms as compared to hundreds of mohms of the ultrafast diode). Accordingly, while the charge of the bootstrap capacitor (which is of hundreds of nF) is almost instantaneous with the diode, longer times (of tens of μs) occur with the integrated diode.
For this reason, it is usual that the first turning on of the low-side MOSFET in the control devices of half-bridges converters with integrated bootstrap diodes is intentionally longer than the following ones during the first switching cycles.
During the pre-charging cycle of the bootstrap capacitor Cboot, having a duration Tpc, if the resonant capacitor Cr is initially charged (this always happens if the split capacitor configuration of Cr is used, shown in FIG. 4), the current Ir will circulate in the resonant circuit. Such a current is a sinusoidal wave at the resonant frequency fR=1/TR of the resonant circuit (Cr, Ls), the peak amplitude of which is equal to the voltage across Cr divided by the characteristic impedance of the resonant circuit itself.
If, at the end of the time period Tpc, the low-side MOSFET Q2 works in the third quadrant (i.e., the current passes from the source terminal to the drain terminal), the current will continue to flow through its body diode Db2, even after the MOSFET Q2 turns off. Therefore, after the dead time Td elapses, the high-side MOSFET Q1 is turned on while the body diode of Q2 is conducting, thus stressing the reversed recovery thereof. FIG. 5 shows the waveforms of the signals HSGD, LSGD, the half-bridge voltage VHB, the voltage Vcr at the terminals of capacitor Cr, the current Ir, the current IQ2 flowing through the transistor Q2, and the current Ilp which flows through the inductor Lp.
The low-side MOSFET Q2 will conduct into the third quadrant at the end of the pre-charging time Tpc if the condition
            K      2        ×          T      R        <  Tpc  <                    K        +        1            2        ×          T      R      is met, where K is an odd integer.
Resonance frequency fR=1/TR of the LLC circuit is typically selected based on other considerations, whereby restraining it to the time period Tpc is not generally acceptable.